Did energy costs of walking limit the evolution of a larger human birth canal?

Abstract Human childbirth is uniquely challenging among primates, which has been attributed to evolutionary trade-offs between birthing large-brained infants and bipedal locomotion (‘obstetrical dilemma’). Adapted for efficient upright walking and running, the human pelvis forms a rigid ring that leaves little leeway for parturition and causes a relatively high risk of obstructed labour. Static mechanics suggests that a wider pelvis increases the energy demands of the abductor muscles during locomotion. Recent empirical studies, however, have not observed a significant increase in whole-body locomotor costs in individuals with wider pelves. To investigate this discrepancy, we employed a detailed musculoskeletal model and predictive forward modelling to simulate human gait while systematically varying body dimensions beyond modern human variation. Our results confirm that a wider pelvis substantially increases abductor muscle energy demands and activation (± 10% for ± 20% width change), consistent with predictions based on lever mechanics. However, during dynamic walking, these increased demands are offset by a redistribution of energy costs to other muscles, leading to only minimal increases in whole-body metabolic cost (± 1% for ± 20% width change). In contrast, lower limb length shows a 2.5-fold greater effect on metabolic cost than pelvic width. These findings reconcile static mechanics predictions with whole-body measurements: while abductor muscle costs increase with pelvic width as predicted, these costs are compensated mainly at the organismal level. We propose that if locomotor costs constrained pelvic evolution, they likely operated through subsystem-level mechanisms, such as muscle-specific energetics and fatigue, rather than whole-body energy economy.